Moving object detection is a key step of classification and tracking.For color images, we usually choose the color feature as the testing standards of moving object. Different color space can reflect different image information. Therefore, this paper analyzes the information of moving object in a variety of color space. Compares the background subtraction effect based on Gaussian mixture model algorithm with different color space characteristic values, under complex background. Experiments result show that, in RGB, YCrCb, HSV and Lab color space, the moving object detection algorithm based on Lab color space has the highest true positive rate and lowest false positive rate, which has anti-interference of illumination changing, strongest adaptability and best robustness%视频图像运动目标检测是有效分类和跟踪运动物体的关键步骤.对于彩色图像,大多选择颜色特征作为运动目标的检测标准,而不同的颜色空间反映图像的信息不同.为此,本文分析了运动目标在多种颜色空间下的信息变化情况,以及不同色彩空间下的像素特征对目标检测算法性能的影响；比较了复杂背景环境下,基于不同颜色空间特征值的混合高斯模型算法的背景减除效果,以期为多种典型环境下颜色特征优选和算法优化提供理论基础和实践指导.实验结果表明:在RGB、YCrCb、HSV和Lab四种不同颜色空间中,基于Lab颜色空间的运动目标检测算法具有更高的检测率和更低的误检率,是一种能抗背景光干扰、适应能力较强、鲁棒性较好的检测算法.
展开▼
